Farmland boundary recognition method and device based on agricultural machinery track, equipment and medium

Abstract

The application relates to a field boundary recognition method and device based on agricultural machinery trajectories, equipment and a medium. The method comprises the following steps: obtaining dimensional information contained in a plurality of agricultural machinery trajectory points; obtaining a plurality of work path line segments based on the dimensional information; rotating the work path line segments to a work coordinate system according to a work main heading angle, and extracting a one-dimensional point column formed by the work path line segments in the direction of a target coordinate axis of the work coordinate system; performing autocorrelation analysis on a density distribution histogram constructed by the one-dimensional point column to determine an equivalent width estimation value; clustering the work path line segments based on the work path line segments and the equivalent width estimation value to obtain a plurality of work field region points; constructing a triangulation for each work field region point, eliminating triangles with a circumscribed circle radius greater than a preset threshold, and forming a target field boundary by using the remaining triangles. The entire process of the application is fully automated, the field boundary can be accurately fitted without manual intervention, and the application is suitable for complex field contours with convex and concave shapes.

Description

Technical Field

[0001] This application relates to the field of agricultural machinery trajectory data processing technology, and in particular to methods, devices, equipment, and media for identifying land parcel boundaries based on agricultural machinery trajectories. Background Technology

[0002] In recent years, with the upgrading of the demand for precision agriculture subsidy supervision, agricultural machinery GNSS (Global Navigation Satellite System) trajectory data has become a research hotspot. In the research and practice of agricultural machinery trajectory plot identification, density clustering, rasterization, and buffer merging have been widely adopted and have achieved considerable automation results.

[0003] Researchers typically cluster trajectory points first, then fit the plot outline using a convex hull algorithm, or rasterize the trajectory to extract boundaries. Some methods directly merge these points into polygons using a fixed-width buffer. These methods can quickly provide results in scenarios with regular plot shapes and simple operation types, facilitating subsequent area statistics and subsidy calculations. However, when faced with complex situations such as repeated compaction, field turns, or road shifts, existing processes often require manual parameter fine-tuning, indicating room for improvement in automation. Summary of the Invention

[0004] In view of this, this application provides a method, device, equipment, and medium for identifying land parcel boundaries based on agricultural machinery trajectories, in order to solve the problem that related technologies require manual fine-tuning of parameters when encountering complex situations in the research and practice of agricultural machinery trajectory land parcel identification, resulting in a low degree of automation.

[0005] Firstly, this application provides a method for identifying land parcel boundaries based on agricultural machinery trajectories, the method comprising: Obtain the dimensional information contained in multiple agricultural machinery trajectory points; Multiple job path segments are obtained based on dimensional information; Rotate the operation path segment to the operation coordinate system according to the main operation heading angle, and extract the one-dimensional point sequence formed by the operation path segment in the direction of the target coordinate axis in the operation coordinate system; Autocorrelation analysis was performed on the density distribution histogram constructed from a one-dimensional point array to determine the equivalent width estimate; Based on the estimated values ​​of the operation path segments and equivalent width, the operation path segments are clustered to obtain multiple operation plot area points; For each work area, a triangulation is constructed, triangles with circumscribed radii greater than a preset threshold are removed, and the remaining triangles are used to form the boundary of the target area.

[0006] Furthermore, multiple job path segments are obtained based on dimensional information, including: Based on dimensional information, the heading angle and main heading interval of each agricultural machinery trajectory point are determined, whereby the main heading interval is used to characterize the effectiveness of the agricultural machinery trajectory point; Obtain the trajectory points of the target agricultural machinery that fall within the main heading range; Construct operation path segments based on target agricultural machinery trajectory points.

[0007] Furthermore, based on dimensional information, the heading angle and main heading range of each agricultural machinery trajectory point are determined, including: Obtain the projection plane coordinate sequence composed of dimensional information; Based on the projection plane coordinate sequence and the preset conditions for effective screening of dimensional information, the agricultural machinery trajectory points are screened to obtain candidate agricultural machinery trajectory points; Obtain the coordinates of the candidate agricultural machinery trajectory points, and calculate the heading angle based on the coordinates; The main heading section is constructed based on the heading angle and the direction of the main peak.

[0008] Furthermore, the operation path segment is rotated to the operation coordinate system according to the main operation heading angle, and the one-dimensional point sequence formed by the operation path segment in the target coordinate axis direction of the operation coordinate system is extracted, including: Determine the main heading angle for the operation based on the heading angles of all operation path segments; The original plane coordinate system is rotated based on the main heading angle of the operation to obtain the operation coordinate system. The original plane coordinate system is a two-dimensional plane rectangular coordinate system obtained by projecting the points on the agricultural machinery trajectory. After rotating the operation path segment to the operation coordinate system according to the main heading angle, obtain the coordinates of each agricultural machinery trajectory point in the operation path segment in the operation coordinate system, and extract all coordinate points in the direction of the target coordinate axis to integrate a one-dimensional point column.

[0009] Furthermore, autocorrelation analysis is performed on the density distribution histogram constructed from the one-dimensional point sequence to determine the equivalent width estimate, including: Based on a one-dimensional point array, construct a density distribution histogram along the target coordinate axis. Autocorrelation analysis was performed on the density distribution histogram to obtain the periodic peak spacing and hysteresis value of the density distribution; The equivalent width estimate is determined based on the periodic peak spacing and hysteresis value.

[0010] Furthermore, based on the estimated values ​​of the operation path segments and equivalent width, the operation path segments are clustered to obtain multiple operation site area points, including: The neighborhood radius is determined based on the equivalent width estimate. Based on the midpoint and neighborhood radius of all operation path segments, cluster the operation path segments to obtain the operation site area points.

[0011] Furthermore, triangulation is constructed for each work area, triangles with circumscribed radii greater than a preset threshold are removed, and the remaining triangles constitute the boundary of the target area, including: Obtain a preset triangle containing all work plot area points; For each work area point within the preset triangulation, construct a circumcircle, and sequentially delete the common edges of the triangles containing the current work area point within the circumcircle. Connect the current work area point to all vertices of the triangle containing the current work area point, and delete the triangles associated with the vertices of the preset triangle to obtain the triangulation. The target triangles are obtained by comparing the circumcircle radius of each triangle in the triangulation with a preset threshold. The target triangles are then used to construct the boundary of the target plot. The target triangles are the remaining triangles that are retained.

[0012] Furthermore, after using the remaining triangles to form the boundary of the target plot, the method also includes: Use the edge that appears only once in the remaining triangle as the extraction boundary; The extracted boundaries form an unordered boundary; The edges of the unordered boundary are converted into an adjacency list. Starting from any vertex in the adjacency list, the edges of the unordered boundary are sequentially concatenated through predecessor marking and unvisited neighbor queries until the starting point is returned to form a closed loop, thus obtaining an ordered boundary. This constitutes the ordered target plot boundary formed by the extracted boundaries.

[0013] Furthermore, the method also includes: The work area is obtained based on the estimated values ​​of the work path segments and equivalent widths. Based on the work path segments, the work mileage and work time are obtained; Based on the boundaries of the target plot, the plot area is obtained; The average working speed is obtained based on the working mileage and working time; The subsidy calculation results are obtained based on the operating area, operating mileage, operating time, plot area, and average operating speed.

[0014] Secondly, this application provides a land parcel boundary identification device based on agricultural machinery trajectory, the device comprising: The first acquisition module is used to acquire the dimensional information contained in multiple agricultural machinery trajectory points; The second acquisition module is used to acquire multiple job path segments based on dimensional information; The extraction module is used to rotate the operation path segment to the operation coordinate system according to the main heading angle, and extract the one-dimensional point sequence formed by the operation path segment in the direction of the target coordinate axis in the operation coordinate system; The autocorrelation analysis module is used to perform autocorrelation analysis on the density distribution histogram constructed from a one-dimensional point column to determine the equivalent width estimate. The clustering module is used to cluster the operation path segments based on the estimated equivalent width of the operation path segments to obtain multiple operation plot area points. The elimination module is used to construct triangulations for each work area point, eliminate triangles whose circumscribed circle radius is greater than a preset threshold, and use the remaining triangles to form the boundary of the target area.

[0015] Thirdly, this application provides a computer device, including: a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the land boundary identification method based on agricultural machinery trajectory described in the first aspect or any corresponding embodiment.

[0016] Fourthly, this application provides a computer-readable storage medium storing computer instructions for causing a computer to execute the land boundary identification method based on agricultural machinery trajectory described in the first aspect or any corresponding embodiment.

[0017] In this embodiment, the dimensional information contained in multiple agricultural machinery trajectory points is obtained, and then multiple operation path segments are obtained based on the dimensional information. The operation path segments are then rotated to the operation coordinate system according to the main heading angle, and a one-dimensional point sequence formed by the operation path segments in the target coordinate axis direction of the operation coordinate system is extracted. Autocorrelation analysis is performed on the density distribution histogram constructed from the one-dimensional point sequence to determine the equivalent width estimate. Then, based on the operation path segments and the equivalent width estimate, the operation path segments are clustered to obtain multiple operation plot area points. Triangulation is constructed for each operation plot area point, and triangles with circumcircle radii greater than a preset threshold are removed. The remaining triangles constitute the target plot boundary. Thus, the entire process in this embodiment is fully automated, accurately fitting plot boundaries without manual intervention. It is suitable for complex convex and concave plot contours and solves problems such as excessive plot segmentation, duplicate or omitted plot areas caused by repeated compaction, field turns, and road shifts. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the specific embodiments of this application or the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0019] Figure 1This is a flowchart illustrating the land boundary identification method based on agricultural machinery trajectory according to an embodiment of this application; Figure 2 This is a schematic diagram of the operation path line segment extraction according to an embodiment of this application; Figure 3 This is a schematic diagram of coordinate transformation according to an embodiment of this application; Figure 4 It is a density distribution histogram according to an embodiment of this application; Figure 5 This is a schematic diagram illustrating the determination of the equivalent width estimate based on an embodiment of this application; Figure 6 This is a clustering and land parcel segmentation diagram of the operation path segments according to an embodiment of this application; Figure 7 This is a schematic diagram of a preset triangle according to an embodiment of this application; Figure 8 This is a diagram of the circumcircle constructed after selecting point A according to an embodiment of this application; Figure 9 This is a diagram of the circumcircle constructed after selecting point B according to an embodiment of this application; Figure 10 This is an illustration of deleting the common side of the triangle containing point B within the circumcircle, according to an embodiment of this application. Figure 11 This is a diagram showing the selection of points C and D and the deletion of common edges according to an embodiment of this application; Figure 12 This refers to the deletion of a triangle diagram associated with a preset triangle according to an embodiment of this application; Figure 13 This is a diagram illustrating the triangulation results according to an embodiment of this application; Figure 14 This is a schematic diagram of a discrete point set according to an embodiment of this application; Figure 15 This is a Delaunay triangulation diagram according to an embodiment of this application; Figure 16 This is a diagram illustrating the edge of a target plot formed by target triangles according to an embodiment of this application; Figure 17 It is a land parcel area with disordered boundaries according to the embodiments of this application; Figure 18 This is an illustration of the process after removing duplicate vertices according to an embodiment of this application; Figure 19 This is an ordered target land parcel boundary diagram according to an embodiment of this application; Figure 20 This is a diagram illustrating the final land parcel boundary fitting effect according to an embodiment of this application; Figure 21This is a block diagram of a land parcel boundary identification device based on agricultural machinery trajectory according to an embodiment of this application; Figure 22 This is a schematic diagram of the hardware structure of a computer device according to an embodiment of this application. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] It should be noted that, in the description of this application, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. The terms "first," "second," etc., in this application are used to distinguish similar objects and are not used to describe a specific order or sequence.

[0022] In current research and practice on agricultural machinery trajectory identification, density clustering, rasterization, and buffer merging have been widely adopted and have achieved considerable automation results. However, when faced with complex situations such as repeated compaction, field turns, or road shifts, existing processes often require manual parameter fine-tuning, indicating room for improvement in automation.

[0023] To address the aforementioned issues, according to an embodiment of this application, a method for identifying land parcel boundaries based on agricultural machinery trajectories is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0024] This embodiment provides a method for identifying land parcel boundaries based on agricultural machinery trajectories, such as... Figure 1 As shown, Figure 1 This is a flowchart of a land parcel boundary identification method based on agricultural machinery trajectories according to an embodiment of this application. The method includes the following steps: Step S101: Obtain the dimensional information contained in multiple agricultural machinery trajectory points.

[0025] Optionally, agricultural machinery trajectory points refer to the points currently collected by GNSS positioning during agricultural machinery operation and movement in the field, arranged in chronological order. They are the core data source for plot boundary identification and operation information statistics.

[0026] In the embodiments of this application, each point includes three dimensions of information: latitude and longitude coordinates, recording timestamp, and instantaneous velocity.

[0027] Step S102: Obtain multiple job path segments based on dimensional information.

[0028] Optionally, the agricultural machinery trajectory points are preprocessed based on dimensional information to mark valid and invalid points. Then, the valid points are connected to form multiple operation path segments, separating the field operation path from paths unrelated to turning around, transferring, etc.

[0029] Specifically, step S102 includes: Step a1: Determine the heading angle and main heading interval of each agricultural machinery trajectory point based on the dimensional information, wherein the main heading interval is used to characterize the effectiveness of the agricultural machinery trajectory point; Step a2: Obtain the trajectory points of the target agricultural machinery that fall within the main heading range; Step a3: Construct the operation path segment based on the target agricultural machinery trajectory points.

[0030] Optionally, to avoid affecting subsequent calculations, the coordinates need to be projected into planar coordinates before formal processing, and points with obvious signal drift anomalies or non-operational states need to be marked. For example, when the instantaneous speed is less than 0.2 km / h or greater than 25 km / h, the point is considered to be likely a parking spot or a point on a field road. If the calculated distance between the current point and the previous point is less than 0.2 m or greater than 150 m, then the GNSS likely has experienced a jump, repetition, or signal drift.

[0031] Among these, speeds less than 0.2 km / h or greater than 25 km / h, and distances between adjacent points less than 0.2 m or greater than 150 m are preset conditions for validating dimensional information. This application embodiment filters multiple agricultural machinery trajectory points based on these preset conditions to obtain candidate agricultural machinery trajectory points, and then obtains the heading angle based on the coordinates of the candidate agricultural machinery trajectory points. Specifically: Input the projection plane coordinate sequence of the original trajectory , and Latitude and longitude coordinates For timestamps, Let be the instantaneous velocity. A point is marked as invalid if any of the following conditions are true, otherwise it is marked as valid.

[0032]

[0033]

[0034] in —Preset minimum speed, default 0.2km / h —Preset maximum speed, default 25km / h —Minimum preset distance, default 0.2m —Preset maximum distance, default 150m The resulting point sequence is denoted as All subsequent steps are based on conduct.

[0035] After marking the valid points, the heading angles of the trajectory sequence points formed by all candidate agricultural machinery trajectory points are estimated. Specifically, for each trajectory point, the instantaneous heading angle is calculated using the coordinate information of its adjacent trajectory points through a first-order difference method to determine the vector direction from the current point to the next point. In particular, there is a one-to-one correspondence between the validity of the trajectory points and the heading angles.

[0036]

[0037] for Execution Normalize the heading angle to Interval.

[0038] After obtaining the effective heading angles, a histogram statistical model is constructed based on the effective heading angle sequence to identify the working path of agricultural machinery in the field. Because the reciprocating operation results in a significant 180° symmetrical bimodal distribution of the effective heading angles, a frequency distribution of the effective heading angles can be established. By detecting the peak position, the main working direction (i.e., the main peak direction) can be determined. Let the main peak be... Regarding the heading angle definition

[0039] All headings satisfy The candidate agricultural machinery trajectory points within this main heading interval are called target agricultural machinery trajectory points, and these target agricultural machinery trajectory points then form the operation path segment. It should be noted that when the heading angle of a candidate agricultural machinery trajectory point enters this interval, it is considered to have an operational driving tendency; otherwise, it is considered to be turning at the field or transferring driving.

[0040] Specifically, based on the heading discrimination criteria, the trajectory is semantically segmented to generate a set of operational path segments. Starting from the trajectory's starting point, the heading angle of each candidate agricultural machinery trajectory point is sequentially determined to see if it continuously falls within the main heading range. When the number of consecutive valid points meeting the condition is greater than or equal to four, these points are connected to form an operational path segment; if a heading angle exceeds the main heading range, the segment is truncated. Through this process, the entire trajectory is divided into several candidate operational path segments, such as... Figure 2 As shown. Each line segment records its geometric length and the number of sampling intervals it contains, providing basic data for subsequent clustering algorithms.

[0041] Step S103: Rotate the operation path segment to the operation coordinate system according to the main heading angle, and extract the one-dimensional point sequence formed by the operation path segment in the target coordinate axis direction of the operation coordinate system.

[0042] Optionally, the operation coordinate system is constructed with the main peak direction of the heading angle histogram as the rotation angle (X' axis is parallel to the average operation heading, and Y' axis is perpendicular to the X' axis).

[0043] Rotate all points of the operation path segments to the operation coordinate system, extract the coordinates of each point on the Y' axis (perpendicular to the operation direction), form a one-dimensional point sequence, and realize the dimensionality reduction processing of the two-dimensional trajectory.

[0044] Specifically, step S103 includes: Step b1: Determine the main heading angle for the operation based on the heading angles of all operation path segments; Step b2: Rotate the original plane coordinate system based on the main heading angle of the operation to obtain the operation coordinate system, wherein the original plane coordinate system is a two-dimensional plane rectangular coordinate system obtained by projecting the agricultural machinery trajectory points; Step b3: After rotating the operation path segment to the operation coordinate system according to the main heading angle, obtain the coordinates of each agricultural machinery trajectory point in the operation path segment in the operation coordinate system, and extract all coordinate points in the direction of the target coordinate axis to integrate a one-dimensional point column.

[0045] Optionally, the goal of line segment clustering is to group adjacent work paths of the same plot into one cluster. If the neighborhood radius ε is set too small, it will over-segment a plot; if it is too large, it will include surrounding road points in the cluster. Experience shows that setting the neighborhood radius to 1.2 to 1.5 times the row spacing yields good clustering results. Without prior width information, the work row spacing must be estimated from the trajectory information to provide reliable input for the DBSCAN clustering algorithm. Generally, unless otherwise specified, the work width is assumed to be equal to the work trajectory row spacing.

[0046] When obtaining the equivalent swath width estimate based on the algorithm, it is necessary to first obtain the main heading angle of the operation, and then adjust the original plane coordinate system by rotating the main heading angle of the operation to construct the operation coordinate system (i.e., the X' axis is parallel to the average operation heading, and the Y' axis is perpendicular to the X' axis).

[0047] Specifically, based on the heading angle sequence of all valid trajectory points, a frequency distribution histogram is constructed to filter out the heading angle intervals (main peak directions) with the highest frequency of occurrence. The core direction corresponding to this interval is the main heading angle (denoted as rotation angle). Furthermore, first, we statistically analyze the frequency distribution of all heading angles (which heading angle interval has the most points), find the "main peak" of the histogram, and take the heading angle corresponding to the main heading angle as the main heading angle. Because agricultural machinery operates in a round trip, the heading angle distribution will show a 180° symmetrical double-peak feature, and the main heading angle is the core operating direction with the highest proportion.

[0048] The core of equivalent width estimation is to use the parallelism of the trajectory of agricultural machinery reciprocating operation to reduce the two-dimensional plane trajectory to a one-dimensional point sequence, and then infer the operation width through the periodic characteristics of the point sequence.

[0049] First, let's define the two coordinate systems: (1) Original plane coordinate system This refers to the two-dimensional Cartesian coordinate system obtained after projecting the GNSS trajectory of agricultural machinery (such as UTM projection), and the original plane coordinates of the trajectory points are denoted as... ,in: The x-coordinate of the original planar coordinate system; This represents the ordinate of the original planar coordinate system. This coordinate system is fixed and independent of the direction of agricultural machinery operation.

[0050] (2) Rotated working coordinate system To align the parallel work trajectories with the coordinate axes, a new local coordinate system (i.e., the work coordinate system) needs to be constructed with the average work heading as the rotation axis, where: Parallel to the average operating heading of the agricultural machinery, the original XY coordinate system rotates around the origin. The x-axis after the angle; and The axis is perpendicular, meaning the original XY coordinate system has rotated around the origin. The vertical axis after the angle; rotation angle This represents the principal heading angle, i.e., the direction of the main peak in the heading angle histogram. A coordinate transformation diagram is shown below. Figure 3 As shown.

[0051] For any original trajectory point Its coordinates in the working coordinate system The following rotational transformation is used to obtain:

[0052] in: The center of rotation; It represents the rotation angle.

[0053] Specific implementation steps: Step 1: Determine the rotation direction. First, calculate the heading angles of all work segments to obtain the main peak direction of the heading angle histogram, and use this as the rotation angle. For example: if the initial heading angle of the work line is 45°, then After rotation transformation, the work row will be parallel to axis.

[0054] Step 2: Perform rotation transformations on all points of the work line segment. Iterate through each point in the work line segment. Substituting into the above rotation formula, calculate its position. Coordinates in a coordinate system .

[0055] Step 3: Extract the target coordinate axes The coordinates of the points on the axis form a one-dimensional point sequence. Collect all the points of the operational line segments. Values, forming a one-dimensional point sequence .

[0056] Step S104: Perform autocorrelation analysis on the density distribution histogram constructed from the one-dimensional point array to determine the equivalent width estimate.

[0057] Optionally, after obtaining the one-dimensional point sequence, a density distribution histogram along the target coordinate axis is constructed based on the one-dimensional point sequence.

[0058] Specifically, following step 3 of the above embodiment, the following steps are obtained: Step 4: Construct a fine-box histogram. First, determine... Range of coordinate values and Then, the bin width of the histogram was set to 0.3m, and the number of points in each bin was counted to obtain... Histogram of density distribution in different directions. For example... Figure 4 The density distribution histogram is shown below.

[0059] Step 5: Perform autocorrelation analysis on the density distribution histogram. The purpose of autocorrelation analysis is to find the repetition period of the density distribution, i.e., the spacing between rows. Let the density distribution of the histogram be... Its autocorrelation function Defined as:

[0060] in, It is a lagged value, that is Offset on the axis; express Offset The similarity between itself and the previous one.

[0061] Step 6: Determine the equivalent width estimate. Use the first period peak spacing obtained from autocorrelation analysis... If the lag value of the next peak and The difference is less than Then it is believed To prevent noise sidelobes from being separated, only retain As the effective period, the estimated equivalent bandwidth at this time = .like Figure 5 There are 6 operating line segments with a true width of 3.5m and an estimated equivalent width of 3.35m.

[0062] The following example illustrates the implementation process of the autocorrelation function: Assuming to receive Given the values ​​{2.0, 2.1, 2.0, 3.5, 3.6, 3.5, 5.0, 5.1, 5.0}, calculate the autocorrelation function.

[0063] Step 1: Construct a density distribution histogram (First convert the discrete points into a histogram) This is a prerequisite for calculating autocorrelation, and the core of it is "binning statistics": Determine the binning range: Locate Minimum and maximum values: =2.0, =5.1; Set the box width (to reduce the width to ensure that adjacent work rows can be distinguished): set it to 0.1m (i.e., each box covers a 0.1m y-interval). Count the number of points for each box (i.e.) ):

[0064] For ease of calculation, the "box" can be simplified to the midpoint of the interval, i.e., f(2.05)=3, f(3.55)=3, f(5.05)=3, and the rest... corresponding =0.

[0065] Step 2: Determine the offset value traversal range The range of τ is 0 to 1. (After exceeding this range, the offset curves do not overlap), that is, τ∈[0,3.1]m, the step size is consistent with the box width (0.1m), and τ traverses all possible τ: 0,0.1,0.2,…,3.1.

[0066] Step 3: Calculate the autocorrelation value R(τ) one by one. Taking τ=0, τ=1.5m, and τ=3.0m as examples, the calculation process is demonstrated: Example 1: τ=0 (no offset) R(0) = f(2.05) × f(2.05 + 0) + f(3.55) × f(3.55 + 0) + f(5.05) × f(5.05 + 0) = 3 × 3 + 3 × 3 + 3 × 3 = 9 + 9 + 9 = 27 (This is the largest autocorrelation value, i.e., "lag = 0 main peak").

[0067] Example 2: τ = 1.5m (equal to line spacing) R(1.5)=f(2.05)×f(2.05+1.5)+f(3.55)×f(3.55+1.5)+f(5.05)×f(5.05+1.5)=f(2.05)×f(3.55)+f(3.55)×f(5.05)+f(5.05)×f(6.55)=3×3+3×3+3×0=9+9+0=18 (Peak value appears, corresponding to line spacing).

[0068] Example 3: τ = 1.6m (noise sidelobe) R(1.6)=f(2.05)×f(3.65)+f(3.55)×f(5.15)+f(5.05)×f(6.65)=3×0+3×0+3×0=0 (no peak value, belongs to noise).

[0069] Step 4: Generate autocorrelation results and filter peak values. Traverse all Then, we obtain R( The sequence of values ​​for ) is shown in the table below:

[0070] Filtering rules: Remove the main peak at τ=0 (R(0)=27); Find the first non-zero peak value of the positive lag (τ>0): R(τ)=18 (first peak value) when τ=1.5m; Checking adjacent peaks: When τ=1.4m, R(τ)=9, peak spacing=1.5-1.4=0.1m<0.5×1.5=0.75m, which is determined to be a noise sidelobe. Only τ=1.5m is retained as the equivalent width estimate.

[0071] Step S105: Based on the estimated value of the operation path segments and the equivalent width, cluster the operation path segments to obtain multiple operation plot area points.

[0072] Optionally, in the above embodiments, a set of operation line segments L is obtained. work ={L1,L2,…,L m These line segments are essentially centerlines left by agricultural machinery during reciprocating operations in the field, but they have not yet been assigned to specific plots. Traditional DBSCAN density clustering directly clusters the original trajectory points, resulting in a large sample size and susceptibility to local density peaks caused by repeated compaction, easily leading to oversegmentation where a plot is divided into multiple subplots. This application's embodiment upgrades the clustering unit from points to line segments, then abstracts each line segment into its center point, using the midpoint of the line segment to replace the line segment strategy. This method preserves the spatial location information of the line segments while reusing the native DBSCAN algorithm, making it simple and fast to implement. Simultaneously, the midpoint falls on the centerline of the work row, and the density distribution highly overlaps with the central area of ​​the plot, preventing significant shifts in the clustering results. This allows adjacent work rows belonging to the same plot to be clustered into one cluster.

[0073] Specifically, step S105 includes: Step c1: Determine the neighborhood radius based on the equivalent width estimate; Step c2: Based on the midpoint and neighborhood radius of all operation path segments, cluster the operation path segments to obtain the operation plot area points.

[0074] Optionally, the DBSCAN algorithm based on the center point of the line segment is used to automatically divide the potential work area. The specific implementation process is as follows: First, the preprocessed set of valid work line segments L... work ={L1,L2,…,L m Geometric feature extraction is performed, and the geometric center coordinates C of each line segment are determined. i =(x i ,y i ) as clustering samples, constructing the sample space C i ={C1,C2,…,C m At this point, the total number of clustered samples equals the number of line segments.

[0075] In the spatial index construction stage, a two-dimensional KD-Tree data structure is used to build a spatial index for the sample set C, which improves the efficiency of neighborhood query and provides a computational basis for density clustering.

[0076] Regarding parameter configuration, the key parameters of the DBSCAN algorithm are determined based on the characteristics of farmland operations: the neighborhood radius ε is set to 1.3 times the equivalent width estimate w (ε=1.3w); the core point neighborhood contains at least 2 samples, so the minimum number of points MinPts is set to 2.

[0077] The clustering process follows the standard DBSCAN algorithm framework: 1. Perform ε-neighborhood queries using KD-Tree to identify core point sets where the number of samples within a neighborhood is ≥ MinPts; 2. Based on the density reachability principle, expand the clusters of core points, grouping density-connected samples into the same cluster; 3. Process boundary points and noise points, ultimately generating K clusters corresponding to K work plot areas. The results of work path clustering and plot segmentation are as follows: Figure 6 As shown.

[0078] Here, the midpoint of the line segment is used instead of the trajectory point as the clustering unit. This preserves the spatial location characteristics while mitigating the impact of local density peaks caused by repeated compaction, thereby reducing the missegmentation rate of the plots and achieving contour fitting for complex plots such as convex and concave shapes.

[0079] Step S106: Construct triangulation for each work area point, remove triangles with circumscribed circles larger than a preset threshold, and use the remaining triangles to form the boundary of the target area.

[0080] Optionally, here, non-overlapping work site area points (i.e., clusters) are obtained through clustering. For each work site area point, triangulation is first performed to construct Delaunay triangulation, forming a triangular grid covering the sites.

[0081] Calculate the circumcircle radius of each triangle and compare it with a preset threshold (determined based on average velocity, sampling interval, and a tolerance coefficient of 1.2-1.5). Triangles with circumcircle radii greater than the preset threshold are discarded.

[0082] After removing triangles from the triangular mesh whose circumscribed circle radius is greater than the preset threshold, the remaining triangle combinations are used to finally fit the target plot boundary that takes into account both convex and concave details.

[0083] Specifically, step S106 includes: Step d1: Obtain a preset triangle containing all work area points; Step d2: Construct a circumcircle for each work area point within the preset triangulation, and sequentially delete the common edges of the triangles containing the current work area point in the circumcircle. Connect the current work area point to all vertices of the triangle containing the current work area point, and delete the triangles associated with the vertices of the preset triangle to obtain the triangulation. Step d3: Obtain the target triangles whose circumcircle radius of each triangle in the triangulation is less than or equal to the preset threshold, and construct the target plot boundary based on the target triangles, where the target triangles are the remaining triangles that are retained.

[0084] Optionally, (1) construct a Delaunay triangulation, taking four points A, B, C, and D as an example: Step 1: First, create a preset triangle (p1, p2, p3), such as... Figure 7 As shown, this preset triangle should include all points in the point set. Add this preset triangle to the triangle list.

[0085] Step 2: Select point A. Since the circumcircle of the default triangle must contain point A, we can directly connect point A to its vertices, splitting the original triangle into three triangles: (p1, A, p2), (p2, A, p3), and (p3, A, p1), and add them to the triangle list. For example... Figure 8 As shown.

[0086] Step 3: Select point B again, and draw the circumcircle of each triangle in the triangle list. Find the triangles whose circumcircle contains point B, namely (p1, A, p3) and (p2, A, p3). For example... Figure 9 As shown.

[0087] Step 4: Delete the common edge of the triangle containing point B on the circumcircle, i.e., edge (A, p3). Then connect point B to all vertices of the triangle on the circumcircle containing point B. For example... Figure 10 As shown.

[0088] Step 5: Select point C again. Following the same procedure as inserting point B, find the triangle whose circumcircle contains point C, delete the common edges, and then connect the vertices of the triangle. Inserting point D is done in the same way. For example... Figure 11 As shown.

[0089] Step 6: Finally, delete the triangles associated with the preset triangle vertices (p1, p2, p3). For example... Figure 12 As shown.

[0090] Step 7: Triangulation results of final points A, B, C, and D. (e.g.) Figure 13 The triangular mesh shown.

[0091] If the number of points in the work area is large, assume a discrete point set. ,like Figure 14 As shown. For input, construct a two-dimensional Delaunay triangulation, such as Figure 15 As shown.

[0092] (2) Set constraints. For each triangle in the Delaunay mesh, calculate its circumcircle radius r and compare it with a preset threshold alpha. If r > alpha, the triangle is considered to be an external connection or noise segment and is discarded; if r ≤ alpha, it is determined to be within the plot area and is retained. The alpha value is determined based on the velocity-sampling interval, expressed as α = v ×( t) ×γ。 v This represents the average velocity, measured in m / s. t) The average time interval is expressed in seconds (s); γ represents the tolerance coefficient, which can be set to 1.2 to 1.5.

[0093] The triangles with r ≤ alpha are called target triangles, and the region formed by all target triangles is called the final target plot edge, such as... Figure 16 As shown.

[0094] In this embodiment, the dimensional information contained in multiple agricultural machinery trajectory points is obtained, and then multiple operation path segments are obtained based on the dimensional information. The operation path segments are then rotated to the operation coordinate system according to the main heading angle, and a one-dimensional point sequence formed by the operation path segments in the target coordinate axis direction of the operation coordinate system is extracted. Autocorrelation analysis is performed on the density distribution histogram constructed from the one-dimensional point sequence to determine the equivalent width estimate. Then, based on the operation path segments and the equivalent width estimate, the operation path segments are clustered to obtain multiple operation plot area points. Triangulation is constructed for each operation plot area point, and triangles with circumcircle radii greater than a preset threshold are removed. The remaining triangles constitute the target plot boundary. Thus, the entire process in this embodiment is fully automated, accurately fitting plot boundaries without manual intervention. It is suitable for complex convex and concave plot contours and solves problems such as excessive plot segmentation, duplicate or omitted plot areas caused by repeated compaction, field turns, and road shifts.

[0095] As an optional embodiment, after step S106 above, the method further includes: Step e1: Use the edge that appears only once in the remaining triangles as the extraction boundary; Step e2: Construct an unordered boundary based on the extracted boundary; Step e3: Convert each edge of the unordered boundary into an adjacency list, and starting from any vertex of the adjacency list, sequentially concatenate each edge of the unordered boundary through predecessor marking and unvisited neighbor query, until returning to the starting point to form a closed loop, thus obtaining an ordered boundary, which constitutes the ordered target plot boundary composed of the extracted boundaries.

[0096] Optionally, from Figure 16 The boundary of the target plot can be determined, and it is composed of multiple triangles. If the edge that appears only once in the remaining triangles is used as the extraction boundary, then the following can be obtained: Figure 17 The area shown is a block of land with disordered boundaries.

[0097] because Figure 17Since the edges within the plot are unordered, we can perform ordering by constructing an adjacency list and chain tracing to fit an ordered closed-loop plot boundary profile.

[0098] Specifically, assume that an unordered set of boundary edges has been obtained. Each edge ,in , Using planar coordinates. The specific steps for ordering boundary edges are as follows: Step 1: Handle duplicate vertices. First, all vertices need to be unique to avoid duplicate storage. For example... Both on the border In the middle, and also on the border In the middle, iterate through all the vertices of the boundary edges, using the coordinates (x, y) as unique identifiers to construct a vertex set. .

[0099] Step 2: Convert unordered boundary edges into an adjacency list. First, initialize an empty dictionary as an adjacency list, with the key being vertex P and the values ​​being all neighboring vertices of that vertex. Iterate through each unordered boundary edge. ,Will Add to In the neighbor list, Add to The neighbor list of each vertex. Finally, remove duplicate vertices to ensure that there are no duplicate vertices in the neighbor list of each vertex, such as... Figure 18 As shown.

[0100] Convert the edges of the unordered boundary into an adjacency list: P1:[P2,P3] P2:[P1,P3] P3:[P2,P1] Step 3: Chain tracing to order boundary edges. Starting from any vertex, using predecessor marking and unvisited neighbor queries, sequentially concatenate boundary edges until returning to the starting point to form a closed loop. The specific operations are as follows: (1) Initialize tracking parameters Selecting the starting vertex: Randomly select one vertex from the vertex set V as the starting point, start = P_s Initialize the current vertex: current = start Initialize the predecessor vertex: prev=None Initialize an ordered list of boundary points: path=[start] (2) Looping to track neighboring vertices If the current vertex has not returned to the starting point, perform the following operations: Get all neighboring vertices from the adjacency list of the current vertex (current), where neighbors = adj[current]. Traverse the neighbors and find the next neighbor vertex that is not the previous vertex. Add next to the ordered list of points path; Update parameters: prev=current, current=next; The loop stops when current == start.

[0101] Step 4: Generate an ordered closed-loop polygon boundary. Connect the vertices in the ordered point list `path` in sequence to form a closed, ordered boundary for the target plot, such as... Figure 19 As shown.

[0102] It should be noted that the disordered boundary is ordered in order to calculate the area of ​​the polygon using the Shoelace formula.

[0103] The final plot boundary fitting result is as follows Figure 20 As shown.

[0104] The embodiments of this application take into account the details of concave land edges, and the resulting land boundary fitting results avoid the limitation of traditional methods that can only fit convex boundaries.

[0105] As an optional embodiment, the method further includes: Step f1: Based on the estimated value of the work path line segment and the equivalent width, obtain the work area; Step f2: Based on the work path segments, obtain the work mileage and work time; Step f3: Based on the target plot boundary, obtain the plot area; Step f4: Based on the work mileage and work time, obtain the average work speed; Step f5: Based on the operating area, operating mileage, operating time, plot area, and average operating speed, the subsidy calculation result is obtained.

[0106] Optionally, the operation information statistics include plot area, operation area, operation duration, operation time, operation mileage, and average operation speed, providing a standard for operation coverage and efficiency evaluation, and laying a data foundation for route optimization and cost accounting.

[0107] Land parcel area: The area of ​​the target land parcel boundary is calculated using the Shoelace formula and is used for subsidy calculation.

[0108] The working area = the length of the working path segment × the estimated equivalent width.

[0109] Work mileage: The lengths of each work path segment are summed up.

[0110] Task time: The time for each task path segment is summed up.

[0111] Average operating speed: the ratio of operating distance to operating time.

[0112] In subsequent subsidy calculations and practical applications, the aforementioned operating area, operating mileage, operating time, plot area, and average operating speed can be input into the subsidy system for staff to compare and correct the data, thus facilitating accurate subsidy calculation results.

[0113] This application embodiment can output the operating area, operating mileage, operating time, plot area and average operating speed required for subsidy calculation from the original GNSS trajectory with one click without relying on manual post-processing. It can automatically complete the data entry work of the subsidy system, providing a simpler and more robust solution for agricultural subsidy supervision and plot management.

[0114] This embodiment also provides a land parcel boundary identification device based on agricultural machinery trajectory. This device is used to implement the above embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0115] This embodiment provides a land parcel boundary identification device based on agricultural machinery trajectories, such as... Figure 21 As shown, it includes: The first acquisition module 2101 is used to acquire the dimensional information contained in multiple agricultural machinery trajectory points; The second acquisition module 2102 is used to acquire multiple job path segments based on dimension information; Extraction module 2103 is used to rotate the operation path line segment to the operation coordinate system according to the main operation heading angle, and extract the one-dimensional point sequence formed by the operation path line segment in the direction of the target coordinate axis of the operation coordinate system; The autocorrelation analysis module 2104 is used to perform autocorrelation analysis on the density distribution histogram constructed from a one-dimensional point column to determine the equivalent width estimate. Clustering module 2105 is used to cluster the operation path segments based on the operation path line segments and the estimated equivalent width to obtain multiple operation plot area points; The elimination module 2106 is used to construct triangulation for each work area point, eliminate triangles whose circumscribed circle radius is greater than a preset threshold, and use the remaining triangles to form the boundary of the target area.

[0116] In this embodiment, the dimensional information contained in multiple agricultural machinery trajectory points is obtained, and then multiple operation path segments are obtained based on the dimensional information. The operation path segments are then rotated to the operation coordinate system according to the main heading angle, and a one-dimensional point sequence formed by the operation path segments in the target coordinate axis direction of the operation coordinate system is extracted. Autocorrelation analysis is performed on the density distribution histogram constructed from the one-dimensional point sequence to determine the equivalent width estimate. Then, based on the operation path segments and the equivalent width estimate, the operation path segments are clustered to obtain multiple operation plot area points. Triangulation is constructed for each operation plot area point, and triangles with circumcircle radii greater than a preset threshold are removed. The remaining triangles constitute the target plot boundary. Thus, the entire process in this embodiment is fully automated, accurately fitting plot boundaries without manual intervention. It is suitable for complex convex and concave plot contours and solves problems such as excessive plot segmentation, duplicate or omitted plot areas caused by repeated compaction, field turns, and road shifts.

[0117] In some optional implementations, the second acquisition module 2102 is used to determine the heading angle and main heading interval of each agricultural machinery trajectory point based on the dimension information, wherein the main heading interval is used to characterize the validity of the agricultural machinery trajectory point; acquire the target agricultural machinery trajectory point that falls within the main heading interval; and construct the operation path segment based on the target agricultural machinery trajectory point.

[0118] In some optional implementations, the second acquisition module 2102 is used to acquire a projection plane coordinate sequence composed of dimensional information; based on the projection plane coordinate sequence and preset conditions for effective screening of dimensional information, the agricultural machinery trajectory points are screened to obtain candidate agricultural machinery trajectory points; the coordinates of the candidate agricultural machinery trajectory points are acquired, and the heading angle is obtained based on the coordinates; the main heading interval is constructed based on the heading angle and the main peak direction.

[0119] In some optional implementations, the extraction module 2103 is used to determine the main heading angle of the operation based on the heading angles of all operation path segments; rotate the original plane coordinate system based on the main heading angle of the operation to obtain the operation coordinate system, wherein the original plane coordinate system is a two-dimensional plane rectangular coordinate system obtained by projecting the agricultural machinery trajectory points; obtain the coordinates of each agricultural machinery trajectory point in the operation coordinate system after the operation path segment is rotated to the operation coordinate system according to the main heading angle of the operation, and extract all coordinate points in the direction of the target coordinate axis to integrate a one-dimensional point sequence.

[0120] In some optional implementations, the autocorrelation analysis module 2104 is used to construct a density distribution histogram along the target coordinate axis based on a one-dimensional point array; perform autocorrelation analysis on the density distribution histogram to obtain the periodic peak spacing and hysteresis value of the density distribution; and determine the equivalent width estimate based on the periodic peak spacing and hysteresis value.

[0121] In some optional implementations, clustering module 2105 is used to determine the neighborhood radius based on the equivalent width estimate; and to cluster the work path segments based on the midpoints and neighborhood radii of all work path segments to obtain work plot area points.

[0122] In some optional implementations, the elimination module 2106 is used to obtain a preset triangle containing all work plot area points; construct a circumcircle for each work plot area point within the preset triangulation, and sequentially delete the common edges of triangles containing the current work plot area point within the circumcircle; connect the current work plot area point to all vertices of the triangle containing the current work plot area point; delete triangles associated with vertices of the preset triangle to obtain a triangulation; obtain target triangles whose circumcircle radius of each triangle within the triangulation is less than or equal to a preset threshold, and construct the target plot boundary based on the target triangles, wherein the target triangles are the remaining triangles that are retained.

[0123] In some alternative embodiments, the device further includes: After constructing the target plot boundary from the remaining triangles, the edge that appears only once in the remaining triangles is used as the extraction boundary. An unordered boundary is constructed based on the extraction boundary. Each edge of the unordered boundary is converted into an adjacency list. Starting from any vertex in the adjacency list, the edges of the unordered boundary are sequentially concatenated through predecessor marking and unvisited neighbor queries until the starting point is returned to form a closed loop, thus obtaining an ordered boundary. This constitutes the ordered target plot boundary constructed from the extraction boundary.

[0124] In some alternative embodiments, the device further includes: Based on the estimated value of the operation path segment and the equivalent width, the operation area is obtained; based on the operation path segment, the operation mileage and operation time are obtained; based on the target plot boundary, the plot area is obtained; based on the operation mileage and operation time, the average operation speed is obtained; based on the operation area, operation mileage, operation time, plot area and average operation speed, the subsidy calculation result is obtained.

[0125] In this embodiment, the land boundary identification device based on agricultural machinery trajectory is presented in the form of a functional unit. Here, a unit refers to an ASIC (Application Specific Integrated Circuit), a processor and memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0126] This application also provides a computer device having the above-described features. Figure 21 The device shown is a land boundary identification device based on agricultural machinery trajectories.

[0127] Please see Figure 22 , Figure 22 This is a schematic diagram of the structure of a computer device provided in an optional embodiment of this application, such as... Figure 22 As shown, the computer device includes one or more processors 10, memory 20, and interfaces for connecting the components, including high-speed interfaces and low-speed interfaces. The components communicate with each other via different buses and can be mounted on a common motherboard or otherwise installed as needed. The processors can process instructions executed within the computer device, including instructions stored in or on memory to display graphical information of a GUI on external input / output devices (such as display devices coupled to the interfaces). In some alternative implementations, multiple processors and / or multiple buses can be used with multiple memories and multiple memory modules, if desired. Similarly, multiple computer devices can be connected, each providing some of the necessary operations (e.g., as a server array, a group of blade servers, or a multiprocessor system). Figure 22 Take a processor 10 as an example.

[0128] Processor 10 may be a central processing unit, a network processor, or a combination thereof. Processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The programmable logic device may be a complex programmable logic device (CAMP), a field-programmable gate array (FPGA), a general-purpose array logic (GDA), or any combination thereof.

[0129] The memory 20 stores instructions executable by at least one processor 10 to cause at least one processor 10 to perform the method shown in the above embodiments.

[0130] The memory 20 may include a program storage area and a data storage area. The program storage area may store the operating system and applications required for at least one function; the data storage area may store data created based on the use of the computer device. Furthermore, the memory 20 may include high-speed random access memory and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some alternative embodiments, the memory 20 may optionally include memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0131] The memory 20 may include volatile memory, such as random access memory; the memory may also include non-volatile memory, such as flash memory, hard disk or solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0132] The computer device also includes a communication interface 30 for communicating with other devices or communication networks.

[0133] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code, which, when accessed and executed by the computer, processor, or hardware, implements the methods shown in the above embodiments.

[0134] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for identifying land parcel boundaries based on agricultural machinery trajectories, characterized in that, The method includes: Obtain the dimensional information contained in multiple agricultural machinery trajectory points; Multiple task path segments are obtained based on the aforementioned dimensional information; The operation path segment is rotated to the operation coordinate system according to the main operation heading angle, and the one-dimensional point sequence formed by the operation path segment in the target coordinate axis direction of the operation coordinate system is extracted; Autocorrelation analysis was performed on the density distribution histogram constructed from the one-dimensional point array to determine the equivalent width estimate; Based on the operation path segments and the estimated equivalent width, the operation path segments are clustered to obtain multiple operation plot area points; For each work area, a triangulation is constructed, triangles with circumscribed radii greater than a preset threshold are removed, and the remaining triangles are used to form the boundary of the target area.

2. The method according to claim 1, characterized in that, The process of obtaining multiple work path segments based on the dimensional information includes: Based on the dimensional information, the heading angle and main heading interval of each agricultural machinery trajectory point are determined, wherein the main heading interval is used to characterize the effectiveness of the agricultural machinery trajectory point; Obtain the trajectory points of the target agricultural machinery that fall within the main heading interval; The operation path segment is constructed based on the target agricultural machinery trajectory points.

3. The method according to claim 2, characterized in that, The determination of the heading angle and main heading interval of each agricultural machinery trajectory point based on the dimensional information includes: Obtain the projection plane coordinate sequence composed of the dimensional information; Based on the projection plane coordinate sequence and the preset conditions for effective filtering of the dimensional information, the agricultural machinery trajectory points are filtered to obtain candidate agricultural machinery trajectory points; Obtain the coordinates of the candidate agricultural machinery trajectory points, and obtain the heading angle based on the coordinates; The main heading interval is constructed based on the heading angle and the main peak direction.

4. The method according to claim 1, characterized in that, The step of rotating the operation path segment to the operation coordinate system according to the main operation heading angle, and extracting the one-dimensional point sequence formed by the operation path segment in the target coordinate axis direction of the operation coordinate system includes: Based on the heading angles of all the aforementioned operation path segments, determine the main heading angle for the operation; The original plane coordinate system is rotated based on the main heading angle of the operation to obtain the operation coordinate system, wherein the original plane coordinate system is a two-dimensional plane rectangular coordinate system obtained by projecting the trajectory points of the agricultural machinery; After the operation path segment is rotated to the operation coordinate system according to the main heading angle, the coordinates of each agricultural machinery trajectory point in the operation path segment in the operation coordinate system are obtained, and all coordinate points in the direction of the target coordinate axis are extracted and integrated into the one-dimensional point sequence.

5. The method according to claim 1, characterized in that, The autocorrelation analysis performed on the density distribution histogram constructed from the one-dimensional point array to determine the equivalent width estimate includes: Based on the one-dimensional point array, a density distribution histogram is constructed along the target coordinate axis. Autocorrelation analysis was performed on the density distribution histogram to obtain the periodic peak spacing and hysteresis value of the density distribution; The estimated equivalent width is determined based on the periodic peak spacing and the hysteresis value.

6. The method according to claim 1, characterized in that, The operation path segments are clustered based on the estimated equivalent width and the operation path line segments to obtain multiple operation site area points, including: The neighborhood radius is determined based on the equivalent width estimate. Based on the midpoints of all the operation path segments and the neighborhood radii, the operation path segments are clustered to obtain the operation plot area points.

7. The method according to claim 1, characterized in that, The process of constructing triangulations for each work area point, eliminating triangles with circumscribed radii greater than a preset threshold, and using the remaining triangles to form the target area boundary includes: Obtain a preset triangle containing all the points of the aforementioned work plot area; For each work site area point within the preset triangulation, construct a circumcircle, and sequentially delete the common edges of the triangles containing the current work site area point within the circumcircle. Connect the current work site area point to all vertices of the triangle containing the current work site area point, and delete the triangles associated with the vertices of the preset triangle to obtain the triangulation. The target triangles are obtained whose circumcircle radius is less than or equal to the preset threshold compared with the preset threshold, and the target plot boundary is constructed based on the target triangles, wherein the target triangles are the reserved remaining triangles.

8. The method according to claim 1, characterized in that, After the remaining triangles are retained to form the boundary of the target plot, the method further includes: The edge that appears only once in the remaining triangle is used as the extraction boundary; The extracted boundaries constitute an unordered boundary. Each edge of the unordered boundary is converted into an adjacency list. Starting from any vertex of the adjacency list, each edge of the unordered boundary is sequentially concatenated through predecessor marking and unvisited neighbor queries until the starting point is returned to form a closed loop, thus obtaining an ordered boundary, which constitutes the ordered target plot boundary formed by the extracted boundary.

9. The method according to claim 1 or 8, characterized in that, The method further includes: Based on the work path segments and the estimated equivalent width, the work area is obtained; Based on the described work path segments, the work mileage and work time are obtained; Based on the target land parcel boundary, the land parcel area is obtained; The average working speed is obtained based on the working mileage and the working time; The subsidy calculation result is obtained based on the operating area, the operating mileage, the operating time, the land area, and the average operating speed.

10. A land parcel boundary identification device based on agricultural machinery trajectory, characterized in that, The device includes: The first acquisition module is used to acquire the dimensional information contained in multiple agricultural machinery trajectory points; The second acquisition module is used to acquire multiple work path segments based on the dimensional information; The extraction module is used to rotate the operation path segment to the operation coordinate system according to the main operation heading angle, and extract the one-dimensional point sequence formed by the operation path segment in the target coordinate axis direction of the operation coordinate system; The autocorrelation analysis module is used to perform autocorrelation analysis on the density distribution histogram constructed by the one-dimensional point column to determine the equivalent width estimate. The clustering module is used to cluster the operation path segments based on the operation path segments and the estimated equivalent width to obtain multiple operation plot area points; The elimination module is used to construct triangulations for each work area point, eliminate triangles whose circumscribed circle radius is greater than a preset threshold, and use the remaining triangles to form the boundary of the target area.

11. A computer device, characterized in that, include: The system includes a memory and a processor, which are interconnected. The memory stores computer instructions, and the processor executes the computer instructions to perform the land boundary identification method based on agricultural machinery trajectories as described in any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the land boundary identification method based on agricultural machinery trajectory as described in any one of claims 1 to 9.

Images